Related Experiment Video
Updated: Sep 20, 2025
11:02
Detecting Somatic Genetic Alterations in Tumor Specimens by Exon Capture and Massively Parallel Sequencing
Published on: October 18, 2013
19.5K
SAVANA: reliable analysis of somatic structural variants and copy number aberrations using long-read sequencing
Hillary Elrick1, Carolin M Sauer1, Jose Espejo Valle-Inclan1
1European Molecular Biology Laboratory, European Bioinformatics Institute, Hinxton, Cambridge, UK.
Nature Methods
|May 28, 2025
Summary
SAVANA accurately detects cancer
Area of Science:
- Genomics
- Cancer Research
- Bioinformatics
Background:
- Accurate detection of somatic structural variants (SVs) and somatic copy number aberrations (SCNAs) is crucial for understanding cancer evolution.
- Current methods face challenges in precision and resolution.
Purpose of the Study:
- Introduce SAVANA, a novel algorithm for detecting somatic SVs and SCNAs.
- Enable single-haplotype resolution detection using long-read sequencing data.
- Estimate tumor purity and ploidy with or without germline controls.
Main Methods:
- Developed SAVANA algorithm for SV and SCNA detection.
- Established benchmarking protocols using replication and read-backed phasing.
- Analyzed 99 human tumor-normal pairs with matched Illumina and nanopore whole-genome sequencing data.
Main Results:
- SAVANA achieves single-haplotype resolution for SV and SCNA detection.
- Demonstrated significantly higher sensitivity and specificity compared to existing algorithms.
- Showed high consistency between long-read (SAVANA) and short-read sequencing results.
Conclusions:
- SAVANA reliably detects somatic SVs and SCNAs using long-read sequencing.
- Establishes best practices for benchmarking SV detection algorithms.
- Facilitates advanced cancer genomics research through improved variant detection.
